preprocessor.py

import re
import pandas as pd

def preprocess(data):
    # Regex pattern for WhatsApp timestamp (capture only the datetime part)
    datetime_pattern = r'(\d{1,2}/\d{1,2}/\d{2,4},\s\d{1,2}:\d{2}(?:\u202f|\s)?[apAP][mM])\s-\s'
    full_pattern = r'\d{1,2}/\d{1,2}/\d{2,4},\s\d{1,2}:\d{2}(?:\u202f|\s)?[apAP][mM]\s-\s'
    
    messages = re.split(full_pattern, data)[1:]
    dates = re.findall(datetime_pattern, data)  # This captures only the datetime part
    
    # Normalize spaces in datetime strings
    clean_dates = []
    for d in dates:
        # Normalize: Add space between time and am/pm if needed
        d_clean = re.sub(r'(\d{1,2}:\d{2})(?:\u202f|\s)?([apAP][mM])', r'\1 \2', d)
        clean_dates.append(d_clean.strip())
    
    df = pd.DataFrame({'user_message': messages, 'message_date': clean_dates})
    
    # Use format='mixed' to auto-detect datetime pattern
    df['message_date'] = pd.to_datetime(df['message_date'], format='mixed', dayfirst=True)
    df.rename(columns={'message_date': 'date'}, inplace=True)
    
    users = []
    messages = []
    for message in df['user_message']:
        if ':' in message:
            entry = re.split(r'([\w\W]+?):\s', message, maxsplit=1)
            if len(entry) >= 3:
                users.append(entry[1])
                messages.append(entry[2])
            else:
                users.append("group_notification")
                messages.append(message)
        else:
            users.append("group_notification")
            messages.append(message)
    
    df['user'] = users
    df['message'] = messages
    df.drop(columns=['user_message'], inplace=True)
    
    # Date parts
    df['only_date'] = df['date'].dt.date
    df['year'] = df['date'].dt.year
    df['month_num'] = df['date'].dt.month
    df['month'] = df['date'].dt.month_name()
    df['day'] = df['date'].dt.day
    df['day_name'] = df['date'].dt.day_name()
    df['hour'] = df['date'].dt.hour
    df['minute'] = df['date'].dt.minute
    
    # Period column
    period = []
    for hour in df['hour']:
        next_hour = (hour + 1) % 24
        period.append(f"{hour:02d}-{next_hour:02d}")
    df['period'] = period
    
    return df